Fish out of water, need help!

[cweb85]

Hi, I'm an inorganic synthetic chemist who has been put in charge of our broken HP 5890 II with FID and 5m capillary column and split/splitless injector. I would really appreciate help on a few things as I've no experience with electronics, engineering, or analytical chemistry!

1. After much work, I've got all the leaky pipes fixed or replaced. I ended up removing the SMART INDICATING REGULATOR cartridges that remove water from the H2 and air lines for the FID. They had massive leaks and the Airgas rep said they weren't needed for the FID lines, especially if I was using UHP grade gas. Confirmation? The rep knew the previous chemist (who was pretty bad at this job) and thought he was just used to seeing them at his old job in Texas and so bought them for those two lines.

2. I believe the ignitor coil is weak or near broken. How brightly should this be glowing? Yesterday (the first time I had things fixed enough to try ignition) the coil glowed extremely weakly, and by the end of the day it was even worse. I can light the castle with a lighter. Should I buy a new ignitor?

3. As I said I have extremely little experience doing this. I've noticed that as the ramping program brings up the oven temperature, there is an increase in the signal from the FID, is this normal?

4. I believe I have a broken mass flow controller for the septum purge. I can only detect very minimal flows from the septum purge vent if I have a huge total flow and turn the septum purge all the way up. Under normal operating conditions I detect no flow at all. If I take the pipe out of the "in" port to the controller, which is the last place I can remove before the purge vent, I can measure large flow rates. So I know it is only the restrictor in that last controller that seems to take the flow down to 0.

5. Last question, how difficult would it be to make this two dimensional? i.e. add an MS, or are there cheap ways to add a second column dimension for GCXGC? I run product development and QC at a small oil company in california. This GC is my most advanced piece of equipment so I'm trying to learn as much as possible.

6. Real last question. The makeup gas is not currently hooked up correctly, I would have to buy new pipe and everything. I've read that 10 mL/min for carrier gas (I'm using N2 but will switch to H2 or He soon) is the minimum for correct FID function. But I'm using a capillary column that's 5m short! So hexane comes out in about 1 second with my current 10/40/400 gas flow setup. Now I'm studying oils and additives, nothing here is volatile at all (Otherwise it would boil out quick during use). Does anyone think I'll have problems running such a high flow rate through such a short column. Most additives have at the very least 100*C boiling points.

7. Real Real last question. I really am a fish out of water here. Can anyone let me know what sort of solvents can't be used with a polydimethylsiloxane column? I'm using hexane right now but I want some more polar solvents. For instance, I have water soluble additives that I want to study, but since I can't use water, I'm not sure what solvent would be appropriate.

Thanks in advance, I really appreciate help in any of these.

[rb6banjo]

Wow! Here goes:

1) If you use grades of H2 and Air that are clean, you don't need additional purification. I run zero grade air and hydrogen. I pass them through a molecular sieve and a carbon filter to trap whatever might be there but my traps are very old. It's probably overkill for me. If your gases are dirty, you'll have lots of noise in your baseline and it will make your detection limit(s) larger.

2) If the ignitor is glowing, it's probably ok. When they go bad the filament breaks and no current can flow though it. Set your gas flows at H2:Air 30:300. If it doesn't light on its own when you press the ignitor button, gently blow air over the chimney. That nearly always works for me.

3) Yes. Most columns bleed a little at larger oven temperatures.

4) Don't worry about the septum purge at this point. Generally, the flow through that port is always quite low and you have "bigger fish to fry" at this point getting your instrument operational.

5) GCXGC is too much for this instrument. It'll cost you way too much to retrofit it. Agilent (Hewlett Packard) stopped supporting this instrument long ago. Finding a MS that will work with it will be a tough task. If you're serious about GCXGC look at what Leco has to offer.

6) What are your column dimensions? Usually, they are defined by a length, a diameter, and a film thickness, e.g., 30 m (long) x 0.32 mm (diameter) x 0.25 µm (film). For a column of this dimension, the optimum helium head pressure is about 7 psig. At agilent.com, you can download their free flow calculator. You can also get it for iphone in the iTunes app store.

7) Generally, you'd like to have a solvent that is compatible with your stationary phase. This will make your peak shapes generally better. When the sample vaporizes in the injector, the carrier gas sweeps it onto the column where it condenses. If your solvent smears out at the head of your column, it will tend to smear out your analytes too. If the solvent condenses nicely at the head of the column, the analytes will dissolve it rapidly as well. You might be able to live with what happens with a polar solvent on a pdms phase but your chromatography will be better if you can stay with a nonpolar solvent.

Good luck.

[Peter Apps]

Just to add to what rb6banjo said.

Unless you have a megabore column (530 um internal diameter) 10 ml/min is way too high a flow rate through the column. With the column only 5 m long you are going to get very little separation anyway, so you cannot afford to lose any by running the gas so fast. For a 0.25 mm id column you need about 0.5 ml/min and for a 0.32 mm i.d. column you need about 1 ml/min. Double this for helium and about 3 - 4 times for hydrogen. Depending what you need to analyse you will probably have to get a longer column. The lack of makeup gas is a minor problem at this stage (as long as the column is installed to the detector with the column end up inside the tube at the bottom of the flame tip - which with a megabore column can cause problems with carrier flow).

Don't even think about going two dimensional until you can do one dimensional with your eyes shut !

Peter

[Consumer Products Guy]

Fish out of water, need help!

[osp001]

5. Last question, how difficult would it be to make this two dimensional? i.e. add an MS, or are there cheap ways to add a second column dimension for GCXGC? I run product development and QC at a small oil company in california. This GC is my most advanced piece of equipment so I'm trying to learn as much as possible.

I've seen 5890/5970 mass specs for as little as $6,000 as late- computer, GC, MS, and autosampler, supposedly guaranteed working. The problem is they're antiques at this point (I know- I'm trying to press one into service just right now), and without third party widgets the latest OS they'll run is W2000. Some companies still support the 5970, and parts are available and reasonably inexpensive by MS standards. Scientific Instrument Services sells all kinds of parts including stuff that's better than OEM, and you can get more parts on eBay and through 3rd party vendors. But if you're still grappling with the bare bones of GC, a 5970 would increase your migraine pain exponentially. Can't even find the manuals these days.

On the plus side, the 5970 is robust by mass spec standards; while the latest generation of mass specs are pretty amazing, the older HPs are kinda bare bones, but they still give very good numbers by any metric. Now, as for taking your *existing* 5890 and coupling it with MS- don't. Just a recommendation.

[cweb85]

I'll just forget about 2 dimensional anything for now, after a day of just trying to run hexane with 2% Octane I can see how quickly this could get frustrating. I can run such a solution and still get octane to elute at 60*C oven in a few minutes? I figured it would practically never elute at such a low temperature, but blank hexane runs confirms it is octane. And so I find I'm barely getting any separation (~20 seconds) between Hexane (60*C), and Octane (125*C). This cannot be correct and is just me barely knowing what I'm doing. But every journey begins with a step I'll learn, I just have many other projects to work on and this would help me greatly with a number of them.

Also, my dimensions are 5 m. 0.53 mm. and df 0.10 μm.


Thanks for all the replies. I appreciate them.

[Consumer Products Guy]

That's a short wide column. Keep head pressure low, like 1 psi, and temperature 40C

[Peter Apps]

Short, wide and with a very thin film. Not what you want for separating octane from hexane. Unless you have a specific application for high boilers you need a different column - which kind exactly depends on what you hope to do with it.

Peter

[muGC]

What is the purpose of your analysis? Select your column based off of that answer.

If you're working for a small oil company, then you probably should be using ASTM methods for most of your tests. The ASTM GC methods list out what GC conditions should be used.

BTW, you have a high temperature simulated distillation (ASTM D7169) column installed. It's designed to elute all the way up to C100. The resolution for this column is poor, especially below C9.

[cweb85]

What is the purpose of your analysis? Select your column based off of that answer.

If you're working for a small oil company, then you probably should be using ASTM methods for most of your tests. The ASTM GC methods list out what GC conditions should be used.

BTW, you have a high temperature simulated distillation (ASTM D7169) column installed. It's designed to elute all the way up to C100. The resolution for this column is poor, especially below C9.

Okay so what is the purpose? Whatever I can do.... The guy who convinced the boss to buy this didn't know what he was doing, and claimed he could easily reverse engineer competitor products and detect metals like Zn and Ca and do elemental analysis for P, N....he made it seem easy. I know better, but I'm hoping for the following.

It would be really great if I could take a cutting oil sample that might contain 80% base oil, and then several percent each of additives like triethanolamine, and resolve them (lets just say 8 ingredients total). Then I really could try to reverse engineer some competitor products. The thing is there is a relatively short list of items people use in this business, so if I can catalogue all our additives, I might be able to see them in competitor products. If I can't even do this, I want to at least be able to use it for QC on my oil products that I sell. But I'm not familiar with the relationship between boiling point/oven temp/ retention time. I would have thought I could easily separate hexane and octane on that column based on boiling point, not relying on phase interaction. But the previous poster said this column won't easily separate them...why? Octane should stay condensed at 60*C oven temp (boiling point 125*C). Even with the injector at 135*C, I slowed down column flow to 1 mL/min and still didn't get great separation of hexane/octane.

[Peter Apps]

To put not too fine a point on it your boss has been lied to.

Since you are now stuck with the problem you probably need to study up on the basics of GC, and have a look at what other people in the oil industry use it for, what it can do, and what it can't.

One of your major problems will be that you not only have to separate the additives from one another, but also from all the hydrocarbons in the oils - the GC has no way of "knowing" what you are looking for and producing peaks only for those components. An MS would help to give you specificity, but your GC is way past its sell by date for linking up with currently available hardware and software.

Peter

[cweb85]

To put not too fine a point on it your boss has been lied to.

Since you are now stuck with the problem you probably need to study up on the basics of GC, and have a look at what other people in the oil industry use it for, what it can do, and what it can't.

One of your major problems will be that you not only have to separate the additives from one another, but also from all the hydrocarbons in the oils - the GC has no way of "knowing" what you are looking for and producing peaks only for those components. An MS would help to give you specificity, but your GC is way past its sell by date for linking up with currently available hardware and software.

Peter

I know it's a pretty bad situation. The boss paid 9K for the setup too....

I've been trying to study as much as I can on it. I can send out samples for MS or AAS when necessary, but if I could just get some of the basic additives analysed it would be a pretty good start. I'll look at some ASTM methods like people suggest to see what sort of columns they use.

[CE Instruments]

Ouch , even $9k is far too much for that old set up !

GC does not rely on boiling point to elute compounds , rather a significant vapour pressure. Many compounds have a significant vapour pressure once they are within 100C of their boiling point This helps GC deal with high MW oils that may have boiling points way above the 400C maximum a GC can use.

Either get a usable ASTM method set up that is useful to you or get a longer column ~50m with a thicker film 1.0um. This would allow you a chance to see the profle of your oils and maybe pick out high concentration additives.

Sorry but you've been sold a dud, as you know to get proper info. out about your oil and to identify the components you either need a GC TOF with GC x GC capability or several GCs along with other techniques.

If you want a GC TOF then a Dani (US Gentech) will be less expensive

[cweb85]

Ouch , even $9k is far too much for that old set up !

GC does not rely on boiling point to elute compounds , rather a significant vapour pressure. Many compounds have a significant vapour pressure once they are within 100C of their boiling point This helps GC deal with high MW oils that may have boiling points way above the 400C maximum a GC can use.

Either get a usable ASTM method set up that is useful to you or get a longer column ~50m with a thicker film 1.0um. This would allow you a chance to see the profle of your oils and maybe pick out high concentration additives.

Sorry but you've been sold a dud, as you know to get proper info. out about your oil and to identify the components you either need a GC TOF with GC x GC capability or several GCs along with other techniques.

If you want a GC TOF then a Dani (US Gentech) will be less expensive

less expensive...lol...you have no idea who I work for. He is unreasonably "cost-conscious" to a fault. I have no idea how that guy convinced him to spend $9K. If I had some cash the next thing we need to buy is an IR.

And thanks for the vapor pressure lesson, that clears up some things for me!

[MSCHemist]

It is possible to hook a 5890 II up to a 5973 GPIB MSD or a 5972 and use chemstation msd B.02 with Windows 2000. You need to have the correct transferline/heater interface also the wiring can be a bit tricky with the old style wire in crimped into black plastic boxes. With enough research and know how I'm sure it could be done.